The field of laser-solid interactions is now quite mature. The most ambitious undertaking of this kind is the thermonuclear-fusion project where usually lithium deutride or D.sup.2 liquid or solid is driven by powerful UV excimer lasers. It has been very difficult to make the process of controlled fusion work; however, some hints of partial success have been reported.
Interaction of laser light with surfaces of more mundane solids can induce crystal phase transformations because of the high temperatures and pressure generated by shock waves. In addition to the p-t shock wave a gaseous plasma is often created at the solid-gas interface. In the vast majority of studies only a single wavelength laser is employed. If such a plasma is formed by multiplexed lasers, ranging in wavelength from UV to IR, it can obviously create a wide variety of chemically reactive species as well as very hot electrons. Such an environment is a unique set of conditions for crystal growth by condensation from the plasma.
Pulse Method--Radiation heating first from the sun, then from other energy sources for crystal growth purposes became popular in the 1960's. Focusing radiation from xenon lamps and CO.sub.2 laser became new research tools for pulling crystals. Meanwhile, the pyrolysis of hydrocarbon gases over diamond surfaces led to codeposition of diamond and graphite and then to what has become the CVD process for diamond growth. The history of the struggle for the elimination of graphite inclusions is described in several papers, including: Angus J. C., Proceedings of the international School of Physics "Enrico Fermi" Course CXXXV, A. Paoletti and A. Tucciarone (Eds.), IOS Press, Amsterdam (1997), p. 1, 9; Badzian A. R. and DeVries R. C., Mater. Res. Bull., 23 (1988) 385; and DeVries R. C., Badzian A. R. and Roy R., Materi. Res. Soc. Bull., 21 (1996) 65. In the paper by DeVries et al., the photothermal approach to diamond synthesis is mentioned very briefly.
As early as 1967, Derjaguin and Fedoseev proposed a pulsed method for diamond growth. (See Derjaguin B. V. and Fedoseev D. V., Surf. Coat. Technol., 38 (1989) 234) Light from a xenon lamp was focused on a diamond substrate to heat the surface. The heating and cooling were achieved by introduction of a chopper with windows to form pulses of light for heating and intervals for cooling. Pulses of heating produced periodic supersaturation of growth species in the vapor over the substrate which then deposited on cooling. On the basis of nucleation theory developed for diamond and graphite growth it has been suggested that both diamond (two dimensional) and graphite (three dimensional) nuclei are formed during the pulse, but graphite nuclei will have a tendency to disappear during the cooling interval because it will be more difficult for graphite nuclei to reach the critical radius. In this way, the selectivity for diamond growth was achieved. Isometric crystals (10-30 .mu.m) and films (10-15 .mu.m) were grown by this method. Process parameters were as follows: pulse duration 10.sup.-2 s, frequency 5 Hz, average temperature 950.degree. C. and H.sub.2 pressure 25 Torr. The temperature reached during pulses was around 1700.degree. C.
Feasibility of high temperature growth has also been demonstrated with CO.sub.2 laser (40 W) and NaCl optics. The deposition of diamond film from CH.sub.4 at pressures of 1-10 Torr was conducted on diamond substrates. Diamond growth was conducted in the range 1100-1600.degree. C. Only graphitic deposits were obtained for C.sub.2 H.sub.2 and C.sub.2 H.sub.4 gases. (See Fedoseev D. V., Varshavskaya I. G., Lavrent'ev A. V. and Derjaguin B. V., Dokl. Akad. Nauk SSSR (Chemistry), 4 (1977) 928.) In another set of experiments, a chopper was used to form pulses of radiation coming from a 10.6 .mu.m CO.sub.2 laser. (See Varshavskaya I. G. and Lavrent'ev A. V., Arch. Nauk Material, 7 (1986) 127.) The duration of the pulses was in the range 10.sup.-2 -500 s. Polycrystalline, black diamond films, up to 5 .mu.m thick, were grown on diamond substrates. The diamond structure has been established by RHEED. The CO.sub.2 laser provides surface heating and the gas phase stays cold, but the illumination enhances chemisorption, which in turn affects nucleation.
Homogeneous nucleation of diamond.--This refers to the theory of nucleation of diamond from the gas phase directed experiments on homogeneous nucleation. (See Fedoseev, D. V., Derjaguin B. V., Varshavskaya I. G. and Levrent'ev A. V., Carbon, 21 (1983) 243 and Fedoseev D. V., Derjaguin B. V. and Varshavskaya I. G., Surf. Sci. Technol., 38 (1989)99.) The experiment was conducted in air. Drops of a liquid hydrocarbon fell down from a capillary tip of a burette filled with hexane or octane. A focused CO.sub.2 laser beam struck perpendicularly the falling drops. A spark of plasma is formed in such situations. The drops evaporated and solid particles were formed. A variety of crystal and amorphous phases have been identified by electron diffraction, among them: diamond, .beta.-carbine and graphite. Nucleation of diamond from the gas phase was demonstrated this way. This also supports the speculation for interstellar formation of diamond, which as agglomerates fell down to Earth and exists as carbonado.
2.3. Phase transformation in solids induced by laser beam.--Formation of metastable structures of metals by rapid cooling led Russian researchers to similar experiments with graphite. (See: Fedoseev D. V., Bukhovetz V. L., Varshavskaya I. G., Lavrent'ev A. V. and Derjaguin B. V., Carbon, 21 (1983) 237; Fedoseev D. V., Varshavskaya I. G., Lavrent'ev A. V. and Derjaguin B. V., Powder Technol., 44 (1985) 125; and Fedoseev D. V., in Synthetic Diamond, edited by K. E. Spear and J. P. Dismukes (John Wiley and Sons, Inc., New York, N.Y.) 1994, pp. 41-56.) Positive results have been obtained upon heating carbon black above 2000 K and cooling at a rate of 10' K/s. The combination of heating and rapid cooling induced phase transformations in graphite. This rate of cooling is possible for 0.1 .mu.m particles assuming radiation loss according to the Stefan-Boltzmann law.
The experiments were conducted with a CO.sub.2 laser. Carbon black was poured through the focal point of the beam with a flux density up to 2000 W/cm.sup.2. In an alternative experiment a laser beam was focused on a piece of polycrystalline graphite and cooling was conducted with the help of liquid nitrogen. The yields of these two processes were very low. Untransformed graphite material was etched out by cool air plasma. The residue was analyzed by electron diffraction. Cubic diamond, a- and .beta.-carbine, chaoite and amorphous phases were determined.
In a similar manner the transformation of hexagonal-BN to cubic-BN and the so-called E-phase was demonstrated. Interestingly the phase transformation of a-quartz to coesite and stishovite also takes place. The phase transformation induced by lasers seems to be a general phenomenon. It has been speculated that the displacement of thermally agitated carbon atoms from the equilibrium positions is relatively high at temperatures around 2500 K and compares to the length of both single (1.54 A) and double (1.42 A) bonds. Rapid cooling can quench different metastable forms.
However, it is fair to say that confidence in the described results was not excessive and it is certain that very little notice or follow up occurred by the world wide community. In spite of the discouraging background with respect to studies on the preparation of a diamond coating on a substrate, investigators persisted, and eventually it came to be known that conversion of quartz to coesite and stishovite required p-T and time conditions much more stringent than diamond formation.
With this in mind, investigators embarked on experiments to confirm and extend the Fedoseev and Derjaguin work. (See, for example, Alm M., DebRoy T., Roy R. and Breval E., Appl. Phys. Lett., 53 (1988) 1687 and Breval E., Alm M., DebRoy T. and Roy R., J. Mater. Sci. Lett., 9 (1990) 1071.) In the course of the next few years this was done and the inventors herein were able to make diamond and to get even higher-pressure phases (e.g., PbO.sub.2 and Fe.sub.2 N type structures of SiO.sub.2) and were able in the case of CaCO.sub.3 to prove that melting was not involved. (See Alm M. DebRoy T. and Roy R., J. Am. Ceram. Soc., 73 (1990) 733.) There is no doubt about the high p-T conditions achieved, there is some evidence for melting in some cases.
N. Kikuchi et al. provided further confirmation for the Russian discovery giving a detailed description of experiments on carbon conversion with a 1 kW CO.sub.2 laser conducted in He atmosphere. The Raman spectrum proved the formation of diamond particles. (See Kikuchi N., Ohsawa Y. and Suzuki I., Diamond Relat. Mater., 2 (1993) 190.)
However, in spite of these efforts, excimer laser excitation did not result in diamond formation. Different types of carbon films, but not diamond, were deposited with laser ablation of carbon targets.